Introduction

MRI has been widely used in clinical practice for the diagnosis of various diseases. However, conventional ¹H MRI does not provide direct biochemical and metabolic information about tissues, such as cell integrity, tissue viability, or predictions about possible treatment outcomes¹. ²³Na MRI, on the other hand, offers a quantitative and non-invasive method to provide some of this complementary information. Despite sodium ions having low magnetic resonance sensitivity, which can limit image acquisition time, advancements in magnetic field strength have led to improved image quality and resolution. This makes ²³Na MRI a potential new biomarker for neurodegenerative disease. This study represents a preliminary investigation into the reproducibility of ²³Na MRI quantification.

Methods

Ten volunteers were scanned twice with a 3T MR Scanner (Elition X, Philips Healthcare). A double-tuned ¹H/ ²³Na birdcage head coil (Rapid Biomedical GmbH, Rimpar, Germany) was used for scanning. For sodium MRI, a 3D TFE UTE sequence was scanned in the sagittal orientation in 14:28 min. The sequence parameters were: TR/TE = 17/0.1 ms, FOV = 272×272×180 mm³, acquired resolution = 4.75×4.75×4.75 mm³, reconstructed resolution = 2.43×2.43×2.38 mm³, radial acquisition, matrix = 56×56 with 76 slices, NSA = 1, acquisition bandwidth = 40 Hz, and flip angle = 90. Five phantoms with known sodium concentrations (20, 40 ,60 ,80, and 100 mM) were placed within the FOV to serve as external references for quantification. A 32-channel ¹H head coil was used for the high-spatial-resolution STAGE (strategically acquired gradient echo) sequence ^2,3 with a scan time of 3:11 min Other parameters were: FOV = 256×192×150 mm³, SENSE factor 2, matrix = 384×192 with 112 slices, acquired resolution = 0.67×1×1.34 mm³, reconstructed resolution = 0.67×0.67×1.34 mm³, TR = 27 ms, five echoes, TE first = 2.9 ms, and echo spacing = 4.7 ms.For the quantification of the total sodium concentration, 5 ROIs were drawn within each of the 5 phantoms , with one ROI per slice. Subsequently, linear regression was performed between the mean signal intensities within the 5 phantoms and their concentrations. Subsequently, the signal intensities of the sodium MRI were converted to TSC. Following that, the TSC maps were co-registered to the STAGE images which serve as a reference of the anatomical structures. For each of the two measurements, 5 circular ROIs were manually drawn on 5 different slices of the fused images of STAGE and TSC, for each of the brain regions including the Substantia Nigra (SN) and Putamen, which are important regions for neurodegenerative diseases such as Parkinson’s disease. Subsequently, the mean sodium concentrations across the 5 ROIs were extracted for each brain region.
Statistical analyses were performed using R (version 4.3.2). The intraclass correlation coefficient (ICC) analyses were conducted to assess the degree of agreement between the two measurements. Correlation analysis was also performed, and the difference between two different scans was plotted against their means with their 95% limits of agreement.

Results

The demographics of the participants are shown in Table 1. The concentration of sodium in five phantoms had good consistency. Figure 1 shows in a representative case, the fused images of STAGE and TSC, and the excellent linear correlation (R² = 0.997) between the signal intensities and sodium concentrations in the five phantoms . The calculated sodium concentrations of the five phantoms and two regions of brain (substantia nigra and putamen) for the two measurements were calculated and analyzed (shown in Table 1), showing good consistency and reproducibility. The ICCs were 0.89, 0.71, 0.64, 0.8 and 0.9 for the phantoms with 20, 40, 60, 80 and 100mM sodium concentrations, respectively. And for substantia nigra and putamen, the ICCs were 0.72 and 0.81 respectively. The correlation plot of the two sodium concentration measurements are shown in Figure2.

Discussion and Conclusion

Our study investigated the reproducibility of tissue sodium intensity measured with ²³Na MRI. Our results showed good reproducibility between the two measurements of ²³Na MRI and TSCs in healthy volunteers in both the Substantia Nigra and Putamen regions, which are of particular importance to study neurodegenerative diseases. This indicates ²³Na MRI can be an effective method with a good reproducibility and consistency for evaluating tissue sodium content. To the best of our knowledge, currently there is no reproducibility study published for ²³Na MRI at 3T. Compared to ultra-high field, 3T ²³Na MRI can be readily added to current clinical scans in patients and can provide valuable information about cell function and brain tissue viability in the neurodegenerative diseases⁴.

Acknowledgements

This work was supported, in part, by the National Natural Science Foundation of China (grant number: 82271954, 81971576); Chinese National Science & Technology Pillar Program (grant number: 2022YFC2009900/2022YFC2009905) and the Innovative Research Team of High-level Local Universities in Shanghai.